Google expands AI-powered flood detection and wildfire systems

For the last several years, Google has been using artificial intelligence to develop a system that can predict floods. It has also been working on wildfire tracking tools. Ahead of the COP27 climate conference taking place next week, the company announced that it is expanding those tools.

First, Google says it will offer flood forecasts for river basins in another 18 countries. Those are Brazil, Colombia, Sri Lanka, Burkina Faso, Cameroon, Chad, Democratic Republic of Congo, Ivory Coast, Ghana, Guinea, Malawi, Nigeria, Sierra Leone, Angola, South Sudan, Namibia, Liberia and South Africa. The company previously offered flood warnings to users in India and Bangaldesh with alerts on Android devices and phones that have the Google Search app installed.

Google is also making a tool called Flood Hub available worldwide. Flood Hub displays flood forecasts on a map and shows when and where they might occur with color-coded pins. The company hopes the tool will help people who are directly at risk from the impacts of flooding, and that it will assist organizations and governments in mobilizing their responses.

“This expansion in geographic coverage is possible thanks to our recent breakthroughs in AI-based flood forecasting models, and we’re committed to expanding to more countries,” Yossi Matias, Google vice president of engineering and crisis response lead, wrote in a blog post. Matias noted that catastrophic flood damage affects more than 250 million people every year. Global warming is likely to result in more flooding, which makes detection systems such as the one Google is working on critical.

Using weather forecast data, the company is able to offer flood warnings up to a week in advance, senior staff engineering manager Sella Nevo told The Verge. The AI model previously used water level gauge data, which limited the advance warning window to around 48 hours.

As for wildfires, Matias wrote that Google detects “wildfire boundaries using new AI models based on satellite imagery and [shows] their real-time location in Search and Maps.” The company said last year that it would make its wildfire tracking tool available worldwide. It’s now using machine learning to improve wildfire detection and monitoring. Initially, the improved tracking tools are available in the US, Mexico, Canada and some areas of Australia. The company also uses data from the National Oceanic and Atmospheric Administration and NASA satellites for wildfire tracking.

Matias also touched on some of the other work Google and parent company Alphabet are doing to mitigate climate change, such as an AI-powered system to make traffic lights more efficient and reduce pollution from idling cars. Meanwhile, Mineral, a project housed under Alphabet’s X moonshot division, is attempting to make the global food system more sustainable and productive.